MLB Data Analysis with AI: Ask Baseball Questions in Plain English

Baseball has always been a numbers sport. For over a century, fans and analysts have obsessed over batting averages, ERAs, and on-base percentages. Then came the Statcast era — and suddenly there were 700,000 pitches a season to work with, each one logged with exit velocity, spin rate, launch angle, and GPS coordinates.

The data is everywhere. The friction is in the questions — the ones that require joining Statcast pitch outcomes to FanGraphs season stats to player IDs just to get a single answer. Until now that meant Python, SQL, and a working knowledge of FanGraphs column naming conventions before you could even start. We thought there was a better way — so we built Baseball Analyst: an open-source baseball analytics tool that lets anyone — fan, analyst, or developer — explore MLB data through plain-English questions, no SQL required.

Demo with Claude

What Is Baseball Analyst?

Baseball Analyst is an open-source MLB data analysis tool that connects your AI agent to real baseball data so you can ask questions in plain English and get answers instantly — no SQL, no Python, no data wrangling.

Under the hood, it connects pybaseball — the community library that pulls data from Baseball Savant, FanGraphs, and Baseball Reference — to PlyDB, so your AI agent can query pitch-level Statcast data, season batting and pitching stats, standings, schedules, and a cross-system player ID register, all at once, in response to plain-English questions.

You ask. The AI figures out the SQL. PlyDB runs it against local Parquet files. You get an answer. No warehouse, no ETL, no cloud.

It covers Statcast data from 2008 onwards, with FanGraphs and Baseball Reference stats going back further.

Who is this for? Baseball fans who want deeper analysis than a box score, sabermetrics enthusiasts who are tired of writing one-off scripts, and developers who want a working example of AI-powered sports analytics with PlyDB.

What Makes This Fun

The real joy is asking the questions you’d never bother with if answering them required writing code.

Who’s actually stealing bases efficiently? It’s easy to sort by stolen base totals — but ask the agent to rank players by success percentage with a minimum attempt threshold, then layer in market value, and you start seeing something interesting. Chandler Simpson and Oneil Cruz are delivering elite stolen base production at a fraction of the cost of peers like Bobby Witt Jr. or Trea Turner. The data finds these inefficiencies instantly.

Does raw power actually translate to team value? ISO (isolated power) and wRC+ correlate at r=0.73 across qualified hitters — strong, but far from perfect. Ask the agent to compute the residual and you immediately surface two groups: players whose power compounds into run creation (Ronald Acuña Jr., George Springer — elite OBP amplifies every extra-base hit), and players whose power leaks away (Eugenio Suarez hits 49 home runs but carries a .298 OBP; Jo Adell has genuine pop but a 26% strikeout rate swallows it). One query, and the whole story is visible.

Who actually wins the trade deadline? Using the game-by-game schedule, compare each team’s win percentage and run differential before and after July 31 each season. Which franchises consistently improve after deadline acquisitions, and which ones add payroll without moving the needle?

Pitch mix evolution: How has the league-wide share of four-seam fastballs changed since 2015? Which starters have most dramatically shifted their arsenal — and did the change correlate with better or worse outcomes?

The platoon advantage by pitch type: Does the left-on-left matchup advantage hold equally for curveballs and sliders? Pull wOBA splits by batter/pitcher handedness and pitch type across the Statcast era.

Which parks are most pitcher-friendly right now? Compare barrel rate allowed and average xwOBA when pitchers are at home vs. away over the last three seasons. Which ballparks consistently suppress or inflate batted ball quality beyond what the season stats suggest?

The aging curve: At what age does offensive production (wRC+) peak for position players, and how does that compare to pitching performance (FIP) for starters vs. relievers? Does the Statcast era show a different curve than the pre-launch-angle era?

Prospect to superstar: Using the player ID table to join Statcast and FanGraphs data, identify the fastest rises from debut to peak wRC+ or FIP. Which teams have been best at developing young players into impact contributors within their first three seasons?

The data supports all of it. The AI does the legwork.

Try It Yourself

The repo has everything you need: a data download script, a pre-configured PlyDB setup, and a semantic overlay so your agent understands baseball-specific encodings and table relationships out of the box.

Setup takes a few minutes. The rabbit holes take considerably longer.

Frequently Asked Questions

Do I need to know SQL to use Baseball Analyst? No. Your AI agent handles all the SQL. You ask questions in plain English; the agent translates them into queries, runs them via PlyDB, and returns the answer.

What baseball data is included? Baseball Analyst pulls from three sources via pybaseball: pitch-level Statcast data from Baseball Savant (2008+), advanced batting and pitching statistics from FanGraphs, standings and schedules from Baseball Reference, and the Chadwick Bureau player ID register for cross-dataset joins.

What AI agent do I use with it? Any PlyDB-compatible agent works. The repo is set up for Claude Code, but the PlyDB config and semantic overlay are compatible with any agent that supports tool use.

Does it work without a cloud database or data warehouse? Yes. All data is stored locally as Parquet files and queried directly by PlyDB. No cloud account, no ETL pipeline, no ongoing infrastructure required.

Is there a similar project for other sports or events? Yes — check out F1 Analyst for Formula 1 telemetry data, Olympics Analyst for the 2026 Winter Games medal analysis, and Oscars Analyst for 97 years of Academy Awards history.

Did you know? PlyDB can connect your AI to boring data too!

Whether it’s business data in a dusty Excel sheet or a complex DevOps log in S3, AI can be surprisingly good at making sense of a mess. PlyDB acts as the bridge, letting your AI query across Postgres, MySQL, CSV, Excel, Parquet, Google Sheets, and more — locally or in the cloud.

Open source and free. Give PlyDB a spin!